A Deep Dive Into The Dynamics of Glycol Chillers

Your question brings up several interesting points that could be expanded into deep dives into topics that most homebrewers would not find very useful, so I will avoid spending too much time crawling down these rabbit holes. However, it does help understand this topic by first reviewing how refrigeration works. The modern human interacts with refrigeration on a regular basis at home and in our automobiles with food refrigerators and air conditioners. Both of these systems typically use a single refrigerant to move energy around.

Refrigerants are used in their gaseous and liquid phases in a closed loop known as the refrigeration cycle. The cycle begins by compressing warm, gaseous refrigerant into hot gas using a compressor, blowing off the superheat in a condenser (the part of the system with a fan that blows hot air away from your refrigerator and makes your kitchen warm), passing the high pressure/not-as-hot liquid refrigerant through an expansion valve and directly into the evaporator coil where energy from the environment (as in your refrigerator) is moved into the refrigerant before exiting the system in the form of warm gaseous refrigerant that returns to the compressor. The parts of the refrigeration cycle that are actually cold are immediately after the expansion valve and in the evaporator coil. The rest of the system is actually warm or hot because energy is being moved from the part of the system being cooled, for example the inside of a beer cooler, to the outside environment.

Most industrial systems use a secondary coolant, such as glycol, to transfer the cold to the application, such as a tank of beer, where the beer gives up its heat to the secondary coolant before the coolant returns to the source of the cold. These systems use a heat exchanger to transfer energy from the glycol (secondary refrigerant) to the primary refrigerant being compressed, cooled, and expanded in the refrigeration cycle. Some industrial systems use ammonia either in direct expansion systems where the liquid ammonia is delivered to the tank jacket or in indirect systems that use glycol as a secondary coolant. Due to leak hazards associated with moving liquid ammonia long distances, direct expansion systems are not very common in modern breweries. Aside from explosion and exposure hazards associated with ammonia, this refrigerant is making a real comeback because it does not damage the ozone layer, is a very efficient refrigerant with terrific thermal properties, and is safer today because of improvements in the design of ammonia refrigeration systems. OK, enough of Refrigeration 101!

The first take-away topic that relates to your question is coolant temperature. In practice, breweries only chill glycol as cold as it needs to be an effective coolant. The rate of cooling is inversely proportional to the difference between the coolant and beer temperature; this means that cooling occurs fastest when beer is warm and slows as the beer temperature approaches the coolant temperature. A good rule of thumb is that the tank coolant needs to be at least 6 °F/3 °C cooler than the coldest desired beer temperature; for example, a tank cooled to 30 °F/-1 °C is best served with 24 °F/-4 °C glycol. If the tank coolant is warmer than this, cooling times become longer and the desired beer temperature may never be achieved.

Using much colder glycol seems like a great way to combat the two challenges identified above because colder glycol means faster beer cooling rate and no problem hitting the target temperature. But there are practical problems involved in achieving this. One is economics; most craft-scale breweries use glycol chillers that have insulated liquid surge tanks used to allow for thermal expansion and contraction in the cooling system (the system includes all the lines connected to the tanks that run throughout a brewery). While these tanks are insulated, energy loss to the environment, or heat gain to the cold glycol, increases as the glycol temperature decreases. Since the glycol is being pumped throughout the brewery, this energy loss perpetuates through all of the glycol lines in the system. There are two easy fixes to this problem. One is to increase the insulation thickness of the entire cooling system and the other is to raise the glycol temperature to the maximum of what works. Since the former is expensive, the latter, and more practical solution (warmer glycol), is the most common.

A good rule of thumb is that the tank coolant needs to be at least 6 °F/3 °C cooler than the coldest desired beer temperature . . .

What you are doing at home to cool your secondary coolant is really clever. Your glycol temperature is much colder than what most commercial craft brewers use because your system design is different. Don’t worry about the differences in what you have because your system is cool and it is working for you. The one major challenge your system presents to you is heat transfer rate. Like a commercial glycol system, you are pumping cold glycol in a loop, starting with your reservoir in your cooler, but your system lacks a heat exchanger to quickly move heat energy out of the return liquid into the air of your freezer. This means that your reservoir temperature will rise over time. As long as the temperature leaving your reservoir is below about 27 °F (-3 °C) you should be fine with the secondary coolant temperature.

The other creative part of your design is the use of a timer to add a time delay into your “on/off” control logic. The control system you describe sends an “on” signal to the pump when the beer temperature is greater than the set point and drops the “on” signal when the measured temperature is at some point in the calculated algorithm of the Grainfather controller. Your time delay puts a big pause into this whole control scheme and decreases the cooling rate by limiting the interval when glycol is actually being pumped. Depending on how your Grainfather controller is configured, this extra step may not be required, but it definitely serves a very real purpose in your system. Especially with the very cold glycol you are using.

In simple terms, you have added an accelerator pedal that governs how much heat you can remove from your tank by limiting the “up time” of your tank cooling system. Too much “up time” makes tank cooling a massive challenge because cold, insulated beer tanks warm up very, very slowly. Unlike overshooting in something like a home cooling system where the air temperature drops below the set point and quickly rises because of heat gain to the system, insulated beer tanks do not quickly warm. If the goal is to chill an ale from 68 °F (20 °C) to 50 °F (10 °C) for an intermediate rest, and the tank inadvertently chills to 39 °F (4 °C), your whole cooling schedule has been thrown off track in a major way! The only real downside to your system is that you may have a hard time getting your beer cold enough because of how you are controlling your pump. If you find that you are unable to hit the bottom end of your cooling range, consider changing the delay on your pump. A more common way to control a system is with a PID (proportional-integral derivative) controller to control something like pump speed or valve position in relation to the differences between set point and the measured value. Your system is different in that your “off interval” does not change as differences between the set point temperature and beer temperature (the measured value) become smaller.

Finally, we arrive at frost burn. The most serious concern related to very cold glycol is freezing your beer. And this usually occurs for one of two reasons with the most common being the location of the temperature probe(s) that measures beer temperature and supplies information to your control system. A brief explanation of beer density makes the topic easier to understand; cold beer sinks at temperatures greater than ~37 °F/3 °C and cold beer rises when beer is cooler than ~37 °F/ 3 °C. This inflection point, or the maximum density of beer, is cooler than the maximum density of water (39 °F/ 4 °C) because of the alcohol and unfermented carbohydrates in beer. This is relevant because the location of the temperature probe(s) in relation to the cooling surface on a tank can lead to blind spots.

Consider a tank with a temperature probe located below the cooling surface of a vertical beer fermenter. When this tank is cooled, cold beer sinks to the bottom as the temperature drops, displaces warmer beer, and the displaced beer rises from the center of the tank to the top; this process is textbook convection. The cooling supply will continue operating until the measured beer temperature falls below the set point. Let’s assume the set point is 4 °F/2 °C to keep the cooling on past the inflection point of beer density. When the beer in contact with the cooling jackets drops below 37 °F/ 3 °C the convection currents in the cooling tank begin to change direction and the cold beer begins to rise to the top along the wall of the tank where the cooling jackets are located. Eventually, the flow of liquid is totally reversed where the cold beer rising up the walls displaces warmer beer on the surface down through the center of the tank (incidentally, this also aids in yeast sedimentation/flocculation).

In the system described, this cooling supply remains flowing until the beer located beneath the cooling jacket (i.e., lower on the tank cylinder) and in contact with the temperature probe is cooler than the 34 °F/ 1 °C. Because the maximum density of beer is ~37 °F/ 3 °C and the tank is not stirred, the bottom of the tank remains right around 37 °F/ 3 °C while the beer above it continues to cool due to this stratification effect. In this scenario, ice will eventually form at the top of the tank because the temperature probe is “blind” to what is happening above its location. The extent of freezing is a function of glycol temperature because the freezing point of beer drops as the liquid beer becomes more concentrated. This is a very common design flaw in many commercial beer tanks that is easily corrected by properly positioning the temperature probe(s). Just don’t locate the probe above the beer level!

When the probe is located above the beer level, especially when paired with a cooling band located in the upper section of the tank, a different sort of blind spot occurs; the probe senses gas temperature instead of beer temperature, the input to the controller is always above the set point, and cooling never shuts off. Some tanks, even well-designed small tanks, have multiple cooling jackets and probes that are used for zone control. In the scenario where the top probe is not submerged in beer, the top of the tank freezes the beer around the upper cooling jacket, while the beer located below this zone is controlled normally.

Commercial breweries face very real production challenges associated with temperature stratification when the root-cause is flawed designs because the fixes can be very expensive. The key points to take away from this topic are the following:

• Well-designed tanks can be cooled to 30 °F/-1 °C using 25 °F/-4 °C glycol. Colder glycol increases cooling rate, but may make overshooting at intermediate temperatures more likely, may increase the risk of tank freezing, increases the cost of line insulation for commercial operations, and gains more energy from the environment than less-cold glycol if intermediate surge tanks are not properly insulated.

• Tanks without bottom cooling jackets are always at risk of having tank bottoms that are no cooler than ~37 °F/3 °C because gravity never takes a day off and beer colder than this rises. This is why lakes and ponds do not freeze into ice cubes in the winter, and why the bottom of larger lakes never drops below 39 °F/4 °C.

• Consider adjusting your on/off delay when your beer temperature is less than about 39 °F (4 °C) if you want a deeper chill.

Prior to my current job at BSG supplying raw materials and other key consumables to breweries, I spent 20 years with Mueller, a stainless steel engineering and manufacturing company in the United States, and we designed and built lots of custom vessels for breweries ranging in size from 20 to 600,000 liters during this time. Just establishing that my thoughts about your question are based on real world experience. Cooling challenges are a pretty cool to discuss, but not cool when experienced in a brewery!

Response by Ashton Lewis.